摘要： 针对第二代Bandelets变换，提出一种新的图像压缩算法。在图像剖分子块时，依据父、子目标函数值的计算比较，优化四叉树剖分，并利用最小二乘法改进剖分区域的最优方向搜索，实现图像压缩，既保留图像丰富的边缘及纹理效果，又降低计算复杂度。实验结果表明，该算法相比JPEG2000的峰值信噪比可提高0.1 dB以上。

关键词: 第二代Bandelets变换, 多尺度分析, 四叉树, 目标函数, 图像压缩

Abstract: According to the second generation Bandelets transform, the paper proposes an algorithm which is based on the objective function calculation from father and son to optimize the quadtree construction when the image is subdivided. The principle of the least squares is also applied to the optimal direction search so as to achieve efficient image compression. The edges and texture features are retained, and the computational complexity is also reduced. Experimental results show the peak signal to noise ratio is increased by more than 0.1 dB compared with JPE2000 scheme.

Key words: the second generation Bandelets transform, multiscale analysis, quadtree, objective function, image compression

中图分类号: 

• N945